Automatic carrier guiding system and guiding method

Abstract

The invention discloses an automatic carrier guiding system and a guiding method. The system comprises a guide sensor group, a control unit and an execution unit, the guide sensor group integrates a visual sensor array and an ultrasonic sensor array, and environment information is accurately obtained; the control unit deeply processes the information and generates a dynamic path planning instruction; and the execution unit faithfully executes the instruction and regulates the running of the automatic carrier. According to the guiding method, multivariate data are fused according to specific steps, and an optimal path is intelligently planned. According to the invention, the traditional guide limitation is broken through, high-precision, low-cost and high-adaptability automatic carrier navigation is achieved, and efficient operation of the logistics industry is enabled.

Description

Technical Field

[0001] The present invention relates to the technical field of logistics handling equipment, and in particular to a guiding system and a guiding method for an automatic transport vehicle. Background Art

[0002] With the rapid development of the modern logistics industry, automated guided vehicles (AGVs) play a vital role in warehouses, distribution centers, and other locations. Their guidance systems, as core components, directly determine the efficiency and accuracy of AAGs. Existing guidance systems suffer from numerous drawbacks. For example, magnetic strip navigation relies on ground-mounted magnetic strips, which is not only cumbersome to install but also extremely inconvenient to maintain. Laser navigation, while offering reasonable accuracy, is susceptible to interference from ambient light fluctuations and lacks stability, making it difficult to adapt to complex and changing logistics scenarios. This severely restricts further improvements in AAG performance. Summary of the Invention

[0003] The present invention aims to overcome the difficulties of existing automatic guided vehicle guidance systems, such as cumbersome installation and maintenance, and navigation accuracy that is easily affected by the environment. The invention strives to create an automatic guided vehicle guidance system that is easy to install, highly adaptable, and has ultra-high navigation accuracy, and also develops a scientific and efficient guidance method to match it.

[0004] Automatic guided vehicle guidance system: The system is composed of a guidance sensor group, a control unit and an execution unit.

[0005] Guidance sensor group: This includes visual sensors and ultrasonic sensor arrays. The visual sensors precisely capture environmental image information including pre-set guidance signs, acting as the AMT's "eyes" to maintain constant awareness of surrounding road conditions. The ultrasonic sensor array comprehensively detects multi-dimensional distance information between the AMT and obstacles, acting like a sensitive "antenna" to ensure safe driving.

[0006] The control unit, the "brain" of the entire system, establishes a robust communication link with the guidance sensor group and the actuator unit. Based on image information, it deeply analyzes the spatial coordinates of guidance markers to accurately locate the target direction. Furthermore, it cleverly integrates multi-dimensional distance information to intelligently generate dynamic path planning instructions, paving the way for the automated guided vehicle.

[0007] Execution unit: includes the drive motor group and steering mechanism, faithfully executes the instructions issued by the control unit, flexibly adjusts the driving speed and direction, and ensures that the automatic guided vehicle reaches its target smoothly and accurately.

[0008] Visual sensor details: A high-frame-rate industrial camera is securely mounted on the center axis of the AMT's roof. Its wide 120° field of view fully covers the area in front of the vehicle, ensuring no critical details are missed. Furthermore, an adaptive illumination compensation module ensures image clarity regardless of light levels, laying a solid foundation for subsequent accurate identification of guide signs.

[0009] Highlights of the ultrasonic sensor array: Composed of four transceiver modules meticulously positioned at the four corners of the vehicle, each sensor array features a precisely set 60° detection angle and a flexibly adjustable detection frequency within a 20-50Hz range, easily capturing obstacle distance information. Notably, the detection blind spot is strictly controlled to within 5cm, significantly improving the reliability of close-range obstacle detection and effectively preventing collisions.

[0010] Control unit architecture: It includes an image processing module and a fusion decision module. The image processing module, equipped with an advanced convolutional neural network algorithm, rapidly identifies QR code guide signs in environmental images and accurately calculates their 3D pose data, making the guide sign's identity clear at a glance. The fusion decision module, using an extended Kalman filter algorithm, performs a spatiotemporal fusion of ultrasonic distance information and 3D pose data, intelligently outputting obstacle avoidance path correction parameters to plan the optimal obstacle avoidance route for the automated guided vehicle.

[0011] Dynamic path planning instruction generation logic: When the ultrasonic sensor array detects that the obstacle is approaching 0.5m, the system quickly triggers the secondary deceleration mode, smoothly reducing the vehicle speed to ensure operational safety; if the obstacle distance is further shortened to 0.2m, the emergency braking mechanism is immediately activated, and the detour path is urgently replanned to cleverly avoid the obstacle and ensure the smooth progress of the mission.

[0012] Actuator upgrades: A new wheel speed feedback module collects real-time speed and torque data from the drive motor, providing first-hand information for driving status monitoring. Combined with a PID closed-loop controller, it dynamically adjusts the PWM duty cycle based on wheel speed feedback, precisely controlling the deviation between the actual driving trajectory and the planned path to within 2 cm, achieving centimeter-level tracking accuracy.

[0013] Automated guided vehicle guidance method: Matching the above sophisticated system architecture, the guidance methods are closely linked.

[0014] Step S1: The visual sensor comes into play and quickly acquires the environment image. After distortion correction, it accurately extracts the feature points of the guidance sign to anchor the key reference for subsequent positioning and navigation.

[0015] Step S2: The ultrasonic sensor array generates a heat map of obstacle distance distribution, which visually presents the distribution of surrounding obstacles and provides detailed data support for path planning.

[0016] Step S3: By integrating the feature point coordinates with the heat map data, the optimal driving path is efficiently calculated using the gradient descent method, taking into account both efficiency and safety, making the automated guided vehicle's route intelligent.

[0017] Step S4: Dynamically adjust the differential steering ratio according to the path curvature to give the automated guided vehicle flexible steering performance, achieve smooth trajectory tracking, and smoothly shuttle through the logistics site.

[0018] Guidance Marker Features: Utilizing a pre-coded QR code matrix, the system embeds a position verification code and a path index code. The position verification code provides a precise benchmark for coordinate system calibration, ensuring zero positioning deviation. The path index code is deeply linked to navigation parameters in a pre-set map database, unlocking the automated guided vehicle's "smart navigation" mode, allowing it to precisely follow the map and reach its destination.

[0019] Path Optimization Strategy: When temporary obstacles intrude, the D*Lite algorithm updates path node weights in real time, allowing for rapid route adjustments and flexible response. If the path conflict rate soars above 30%, global path replanning is triggered, comprehensively reshaping the navigation route to ensure efficient mission execution.

[0020] Differential steering ratio calculation: Following a specific formula, the system integrates parameters such as left and right wheel speeds, wheelbase, and path curvature radius to accurately calculate the differential steering ratio, keeping the control error within 0.1 rad. This gives the automated guided vehicle excellent steering accuracy and allows it to easily navigate complex curves.

[0021] Compared with the prior art, the present invention has the following beneficial effects:

[0022] The automated guided vehicle guidance system and method of this invention offer significant advantages. The powerful combination of visual sensors and ultrasonic sensor arrays complements each other, making them immune to light interference and eliminating reliance on magnetic strips, significantly reducing installation and maintenance costs. The intelligent decision-making of the control unit, coupled with precise control by the actuator unit, achieves centimeter-level navigation accuracy, significantly improving the operating efficiency and safety of automated guided vehicles and injecting strong momentum into the modern logistics industry. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0024] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0025] Automatic guided vehicle guidance system: The system is composed of a guidance sensor group, a control unit and an execution unit.

[0026] Guidance sensor group: This includes visual sensors and ultrasonic sensor arrays. The visual sensors precisely capture environmental image information including pre-set guidance signs, acting as the AMT's "eyes" to maintain constant awareness of surrounding road conditions. The ultrasonic sensor array comprehensively detects multi-dimensional distance information between the AMT and obstacles, acting like a sensitive "antenna" to ensure safe driving.

[0027] The control unit, the "brain" of the entire system, establishes a robust communication link with the guidance sensor group and the actuator unit. Based on image information, it deeply analyzes the spatial coordinates of guidance markers to accurately locate the target direction. Furthermore, it cleverly integrates multi-dimensional distance information to intelligently generate dynamic path planning instructions, paving the way for the automated guided vehicle.

[0028] Execution unit: includes the drive motor group and steering mechanism, faithfully executes the instructions issued by the control unit, flexibly adjusts the driving speed and direction, and ensures that the automatic guided vehicle reaches its target smoothly and accurately.

[0029] Visual sensor details: A high-frame-rate industrial camera is securely mounted on the center axis of the AMT's roof. Its wide 120° field of view fully covers the area in front of the vehicle, ensuring no critical details are missed. Furthermore, an adaptive illumination compensation module ensures image clarity regardless of light levels, laying a solid foundation for subsequent accurate identification of guide signs.

[0030] Highlights of the ultrasonic sensor array: Composed of four transceiver modules meticulously positioned at the four corners of the vehicle, each sensor array features a precisely set 60° detection angle and a flexibly adjustable detection frequency within a 20-50Hz range, easily capturing obstacle distance information. Notably, the detection blind spot is strictly controlled to within 5cm, significantly improving the reliability of close-range obstacle detection and effectively preventing collisions.

[0031] Control unit architecture: It includes an image processing module and a fusion decision module. The image processing module, equipped with an advanced convolutional neural network algorithm, rapidly identifies QR code guide signs in environmental images and accurately calculates their 3D pose data, making the guide sign's identity clear at a glance. The fusion decision module, using an extended Kalman filter algorithm, performs a spatiotemporal fusion of ultrasonic distance information and 3D pose data, intelligently outputting obstacle avoidance path correction parameters to plan the optimal obstacle avoidance route for the automated guided vehicle.

[0032] Dynamic path planning instruction generation logic: When the ultrasonic sensor array detects that the obstacle is approaching 0.5m, the system quickly triggers the secondary deceleration mode, smoothly reducing the vehicle speed to ensure operational safety; if the obstacle distance is further shortened to 0.2m, the emergency braking mechanism is immediately activated, and the detour path is urgently replanned to cleverly avoid the obstacle and ensure the smooth progress of the mission.

[0033] Actuator upgrades: A new wheel speed feedback module collects real-time speed and torque data from the drive motor, providing first-hand information for driving status monitoring. Combined with a PID closed-loop controller, it dynamically adjusts the PWM duty cycle based on wheel speed feedback, precisely controlling the deviation between the actual driving trajectory and the planned path to within 2 cm, achieving centimeter-level tracking accuracy.

[0034] Automated guided vehicle guidance method: Matching the above sophisticated system architecture, the guidance methods are closely linked.

[0035] like Figure 1 As shown, step S1: the visual sensor comes into play, quickly acquires the environment image, and after distortion correction, accurately extracts the feature points of the guidance mark to anchor the key reference for subsequent positioning and navigation.

[0036] The guidance sign in step S1 is a pre-coded two-dimensional code matrix, which includes: a position check code for coordinate system calibration; a path index code associated with navigation parameters in a preset map database.

[0037] Step S2: The ultrasonic sensor array generates a heat map of obstacle distance distribution, which visually presents the distribution of surrounding obstacles and provides detailed data support for path planning.

[0038] Step S3: By integrating the feature point coordinates with the heat map data, the optimal driving path is efficiently calculated using the gradient descent method, taking into account both efficiency and safety, making the automated guided vehicle's route intelligent.

[0039] Step S3 also includes: when a temporary obstacle is detected, updating the path node weights in real time based on the D*Lite algorithm; when the path conflict rate is greater than 30%, triggering global path replanning.

[0040] Step S4: Dynamically adjust the differential steering ratio according to the path curvature to give the automated guided vehicle flexible steering performance, achieve smooth trajectory tracking, and smoothly shuttle through the logistics site.

[0041] The calculation formula of the differential steering ratio in step S4 is:

[0042]

[0043] Among them, v L 、v Ris the left and right wheel speed, L is the wheelbase, R is the path curvature radius, and the control error is ≤0.1rad.

[0044] Guidance Marker Features: Utilizing a pre-coded QR code matrix, the system embeds a position verification code and a path index code. The position verification code provides a precise benchmark for coordinate system calibration, ensuring zero positioning deviation. The path index code is deeply linked to navigation parameters in a pre-set map database, unlocking the automated guided vehicle's "smart navigation" mode, allowing it to precisely follow the map and reach its destination.

[0045] Path Optimization Strategy: When temporary obstacles intrude, the D*Lite algorithm updates path node weights in real time, allowing for rapid route adjustments and flexible response. If the path conflict rate soars above 30%, global path replanning is triggered, comprehensively reshaping the navigation route to ensure efficient mission execution.

[0046] Differential steering ratio calculation: Following a specific formula, the system integrates parameters such as left and right wheel speeds, wheelbase, and path curvature radius to accurately calculate the differential steering ratio, keeping the control error within 0.1 rad. This gives the automated guided vehicle excellent steering accuracy and allows it to easily navigate complex curves.

[0047] System construction: First, carefully select an appropriate high-frame rate industrial camera, accurately install it on the central axis of the roof according to the model of the automatic transport vehicle and the requirements of the working scene, debug the field of view to 120°, and complete the calibration of the adaptive light compensation module. Ensure that the camera's field of view covers the key area in front of the vehicle body, without blind spots. Then, evenly arrange four sets of ultrasonic sensor transceiver modules at the four corners of the vehicle body, calibrate the detection angle to 60° one by one, set the detection frequency range to 20-50Hz, and strictly inspect the detection blind area to ensure that it is within 5cm. Next, assemble the control unit, embed the image processing module and the fusion decision module, import the convolutional neural network and extended Kalman filter algorithm, and debug it to the optimal operating state. Finally, assemble the execution unit, install the drive motor group, steering mechanism and wheel speed feedback module, configure the PID closed-loop controller, and complete the system hardware construction.

[0048] Parameter Tuning: For the visual sensor, we collected environmental images under various lighting conditions and repeatedly trained the convolutional neural network using professional image processing software, optimizing the recognition algorithm parameters and improving the accuracy of QR code guide sign recognition. For example, through extensive training with a large number of samples, recognition accuracy was increased to over 99% under the mixed lighting conditions of fluorescent and natural light commonly found in warehouses. For the ultrasonic sensor array, we simulated various obstacle scenarios, fine-tuned the detection frequency, and calibrated the distance measurement accuracy to ensure accurate obstacle distance information. In the control unit, we repeatedly tuned the parameters of the image processing module and the fusion decision module through extensive simulation experiments to ensure the scientific and reasonable dynamic path planning instructions. The execution unit calibrated the wheel speed feedback module based on the dynamic characteristics of the automated guided vehicle and optimized the PID closed-loop controller parameters to achieve precise speed and direction control. For example, we tuned the PID parameters under different operating conditions, including when the guided vehicle was fully loaded and unloaded, to ensure that the driving trajectory deviation was within 2 cm.

[0049] Implementation method: When an automated guided vehicle (AGV) is activated, its visual sensors immediately activate, capturing environmental images at high frequency. After distortion correction, they rapidly extract guide sign feature points and accurately locate its position. For example, when navigating aisles between shelves in a large logistics warehouse, it can accurately identify pre-set guide signs on both sides of the aisle, determining its direction and distance. Simultaneously, an ultrasonic sensor array operates synchronously, detecting obstacle distances in all directions and generating a heat map. For example, when the AGV turns into a narrow aisle, the heat map clearly displays the distances to shelves on both sides of the aisle and obstacles ahead. After the control unit receives image and distance information, the image processing module rapidly identifies the QR code guide signs and calculates the three-dimensional pose data. The fusion decision module uses the extended Kalman filter algorithm to fuse multiple information, output obstacle avoidance path correction parameters, and generate dynamic path planning instructions. Upon receiving the instructions, the execution unit drives the motor group and steering mechanism to work in coordination. Based on data from the wheel speed feedback module, the PID closed-loop controller adjusts the PWM duty cycle in real time, driving the AGV to smoothly travel and accurately track the planned path. If a temporary obstacle is encountered during travel, the system updates the path node weights in real time using the D*Lite algorithm. If the path conflict rate exceeds 30%, global path replanning is immediately triggered to ensure smooth mission execution. For example, if cargo suddenly falls and blocks part of a transport corridor, the system quickly replans the route to bypass the obstacle and continue the transport mission efficiently.

[0050] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. An automatic guided vehicle guidance system, characterized in that: include: It includes a guidance sensor group, a control unit and an execution unit; the guidance sensor group includes a visual sensor and an ultrasonic sensor array, the visual sensor is used to obtain environmental image information containing preset guidance signs, and the ultrasonic sensor array is used to detect multi-dimensional distance information between the automatic guided vehicle and obstacles; the control unit is communicatively connected with the guidance sensor group and the execution unit, and is configured to parse the spatial coordinates of the guidance signs based on the image information, and generate dynamic path planning instructions in combination with the multi-dimensional distance information; the execution unit includes a drive motor group and a steering mechanism, which is used to adjust the driving speed and direction according to the dynamic path planning instructions.

2. The automatic guided vehicle guidance system according to claim 1, characterized in that: The visual sensor is a high-frame-rate industrial camera installed on the top central axis of the automatic transport vehicle. Its field of view covers a 120° range in front of the vehicle body and is equipped with an adaptive light compensation module.

3. The automatic guided vehicle guidance system according to claim 1, characterized in that: The ultrasonic sensor array includes four groups of transceiver modules, which are respectively arranged at the four corners of the vehicle body. The detection angle of each sensor group is 60°, the detection frequency is 20-50Hz, and the detection blind area is ≤5cm.

4. The automatic guided vehicle guidance system according to claim 1, characterized in that: The control unit includes: an image processing module that uses a convolutional neural network to identify the QR code guide signs in the environmental image and calculates their three-dimensional posture data; a fusion decision module that performs spatiotemporal synchronous fusion of ultrasonic distance information and three-dimensional posture data based on the extended Kalman filter algorithm and outputs obstacle avoidance path correction parameters.

5. The automatic guided vehicle guidance system according to claim 4, characterized in that: The generation of the dynamic path planning instruction includes: when the obstacle distance is detected to be ≤0.5m, triggering the secondary deceleration mode; when the obstacle distance is ≤0.2m, activating emergency braking and replanning the detour path.

6. The automatic guided vehicle guidance system according to claim 1, characterized in that: The execution unit also includes: a wheel speed feedback module that collects the speed and torque data of the drive motor group in real time; a PID closed-loop controller that dynamically adjusts the PWM duty cycle based on the wheel speed feedback data to ensure that the deviation between the actual driving trajectory and the planned path is ≤2cm.

7. The method for guiding an automated guided vehicle according to claim 1, wherein: The system according to any one of claims 1 to 6 is characterized in that it includes: S1. acquiring an environmental image through a visual sensor, and extracting characteristic points of a guide sign after distortion correction; S2. generating a heat map of obstacle distance distribution through an ultrasonic sensor array; S3. calculating an optimal driving path using a gradient descent method based on the coordinates of the characteristic points and the heat map data; S4. dynamically adjusting the differential steering ratio according to the path curvature to achieve smooth trajectory tracking.

8. The method according to claim 7, characterized in that The guidance sign in step S1 is a pre-coded two-dimensional code matrix, which includes: a position check code for coordinate system calibration; a path index code associated with navigation parameters in a preset map database.

9. The method according to claim 7, characterized in that Step S3 also includes: when a temporary obstacle is detected, updating the path node weights in real time based on the D*Lite algorithm; when the path conflict rate is greater than 30%, triggering global path replanning.

10. The method according to claim 7, characterized in that The calculation formula of the differential steering ratio in step S4 is: Among them, v L 、v R is the left and right wheel speed, L is the wheelbase, R is the path curvature radius, and the control error is ≤0.1rad.

Images